Hadoop and Hana drive digitization forward

Technologies such as Hana and Hadoop are opening up new opportunities here. The current Lünendonk trend study "Creating added value through digital transformation" shows that many companies are hesitant when it comes to digital transformation and innovative developments.

The study authors recognize a danger in the tendency to classify opportunities arising from the development of new business models or the opening up of new markets as less significant than "defensive goals" such as customer loyalty or process improvements.

The conclusion is that such an assessment makes bold and disruptive innovations in business models, products and services more difficult. But how can companies drive digital transformation and test new business models?

Traditional systems and architectures are not designed to cope with the flood of data and data sources in an agile manner. Big data can therefore only become big business if the right technologies and organizational structures are used and established.

Modern databases use high-quality hardware components to process large amounts of data and process the data in main memory (in-memory).

These technologies are extremely powerful, but also expensive. The flexibility to store large amounts of data at short notice without running up against budget limits and thus increasing the pressure to perform is essential for the digital transformation.

This is where the open source Hadoop platform comes in, because it was created for exactly this purpose: storing, processing and analyzing very large volumes, hundreds of petabytes to zetabytes of data.

A key advantage over other systems is that Hadoop is not based on expensive proprietary hardware for storing and processing the data. The advantage of the distributed file system also extends to the distributed processing of the data and can scale almost endlessly via inexpensive standard servers.

New paths for Big Data Analytics

However, in many cases Hadoop alone is not sufficient for the requirements of Big Data Analytics. For the evaluation of unstructured or semi-structured data in combination with the latest business data, in-memory processing with modern analysis methods is the best solution.

Hana and Hadoop are a strong team: The combination of high-performance database and solid mass data platform can open up new paths for business analytics and convinces with massive cost savings.

Near-Real-Time Analytics

Data from sensors, networks, and machines, as well as unstructured information from photos, texts, and social networks, can be analyzed cost-effectively and in near real time with the help of Hana and Hadoop, enabling countless application scenarios, as the app of an international agricultural company shows.

With the help of this app, farmers can detect plant diseases at an early stage via smartphone or tablet and select the right treatment.

If the farmer sends a photo of the diseased plant, it is immediately compared with reference images of plant diseases in the database using data mining. In addition, current weather data for the region is included in the analysis.

In the background, Hana works out a recommendation in real time, which the farmer then receives immediately on his mobile device. This contains the determination of the plant disease as well as a scheme of how best to treat the disease based on the current weather conditions.

The mass data required for this, such as product data sheets, videos or image reference data, are stored in a Hadoop cluster. A major advantage of Hadoop is data streaming. Mass data can be analyzed directly in Hadoop using statistical models.

Only the essence of these evaluations is then passed on to Hana. This approach makes it possible to evaluate huge amounts of data with different structures in real time and with high performance.

Another advantage is that the combination of Hana and Hadoop is also suitable for smaller companies with limited IT resources. This is because with predictive analytics, SAP offers standard analyses that enable companies with little statistical know-how to generate corresponding models.

Courage to try things out is the key to success here. Datavard offers innovation workshops in which IT managers can learn how to apply new technologies and test ideas and business models at the same time.